Springs in shells
Springs in shells
(OP)
Hello from Spain:
I have calculated a water tank but i´m not sure that the results were good and i think that the problem are the springs that I put in the shells because the strange resuts are at the joint between the wall and the floor. So I have three questions:
1st) What simulate better the soil reaction, springs or Nlinks
2nd) Where can i find an example of a slab (I think that is the name of the floor)
3rd) What are better for shells springs in joins or in shells and what´s the difference.
Thanks in advance.
I have calculated a water tank but i´m not sure that the results were good and i think that the problem are the springs that I put in the shells because the strange resuts are at the joint between the wall and the floor. So I have three questions:
1st) What simulate better the soil reaction, springs or Nlinks
2nd) Where can i find an example of a slab (I think that is the name of the floor)
3rd) What are better for shells springs in joins or in shells and what´s the difference.
Thanks in advance.
Talk To Other Members
RE: Springs in shells
Define an area/shell with the properties of your slab, 3 foot thick concrete, for example. Typically, you would give the area a "shell" type which gives stiffness in-plane and out of plane. If you have version 8, you can use the Edit menu to extrude a shell into solid elements and then assign springs to the solid element if you want. In older versions of SAP, you have to import autocad or edit with the text file in order to model solid elements
For soil, a spring restrains joints in both tension and compression. I hate to keep bringing up version 8 if you don't have it, but version 8 lets you make an area assignment for springs which automatically calculates nodal spring constants for you. This saves you the trouble of manually figuring out joint spring constants based on element size and tributary area. If you have a subgrade modulus of .3 kip-in, you would simply assign>area>area spring and assign an area spring value of .3 kip-in to your mat foundation and the program automatically calculates joint spring constants for you.
If you want to consider the possibility that the foundation may uplift and redistribute load, then you will need to assign links - gap type links, at joints instead of springs. Unfortunately, there is no area assignment for links as there is for springs. You will need to calculate the soil spring constant based on size of element and tributary area
One other bit of advice. You should consider putting soil springs in all three translational directions, not just vertical, to prevent numerical instabilities.
RE: Springs in shells
I put springs only in the face number 6 and the value in in the direction 1 and 2 is 10000 KN/m3 and 20000 in the direction 3.
What´s the problem? without loads the deformation and the stresses should must be zero.
RE: Springs in shells
Did you ignored the own wieght of the mat?
RE: Springs in shells